( 253 ) 



apply tliein to plienomena belonging- to regions for wiiicii their 

 applieability has not been expei-inientcilly proved. A generalization 

 of this kind is of course iiypothetical. The fact tliat the two deductions 

 here yield the same results will probably l)c cojisidered as a conti r- 

 niation of the validity of the iij'potheses. 



The qnestion suggests itself whether the hypothesis concerning the 

 mass of the energy is not only in the special cases treated above, 

 but with [)erfeet generalily in agreement with the theoj-y of relativity. 



The most general method to solve this (piestion seems to be thai 

 suggested by Laue. His argument comes in principle to the following. 

 We will take the 16 quantities 



and differentiate the four quantities of one horizontal row respect i\ely 

 according to ./', //, z arid let and put the sum of the four terms 

 thus obtained equal to zero \). The four horizontal rows yield four 

 equations of this kind; the first three equations determine the increase 

 of the momentum, the fourth equation is an expression of the law 

 of conservation of energy. We have chosen for the elements of the 

 fourth vertical column the same (piantities which occur in the forth 

 horizontal row. By making this choice we have introduced the 

 hypothesis of the mass of the energy. 



IjAUK now postulates that these 'J(i (pumtities, when we make use 

 of a moving coordinate system will be transformed as the elements 

 of a fourdimensional tensor, (in this way the e([uations (102) cited 

 above are found) and so he postulates that t lie hypothesis concerning 

 the mass of the energy agrees with the hypothesis of relativity. The 

 <piestion must however be put: have we a right to [)Ostulate that 

 the quantities will ti-ansform in the way given by Laue? We must 

 keep in view that \ve are dealing with derived c|uantities. From the 

 e([uati()n ^ , z^ 2£ t^AM ,. e.g. it appears that, if we have already assumed 

 in what way (i and a\ will be trausfonucd, the formula for the 



1) The clioico zenj lor llit; righlliand iiieml)cr of tlic t'Cjualions is uii cxpr(_'s.sion 

 of the hypothesis that no ''aclio in liislans" occurs. Laue does not introduce; this 

 hypollicsis, his cqualions llicrerore have a righthand member dilleiing from zero. 



